London

to, heart failure patients, and to identify developments of particular interest or expertise in this area. Method: Postal survey of all lead consultants of English SPC services, September 2004. Results: Of 397 services, 233 replied (response rate 59%); 222 (95%) thought SPC had a role in severe/end stage heart failure, while three (1%) did not. A total of 197 services (85%) accepted heart failure patients, 26 (11%) did not. The most common reasons for not accepting heart failure patients were lack of resources or beds, implications for staff training or an organizational decision. The mean number of heart failure patients currently under a service was 2.2, but 15 had more than five (maximum 53). Fifteen services (6%) had specific referral criteria for heart failure patients, including recurrent hospital admissions without symptomatic improvement, inappropriateness of further hospital admission and severity of heart failure. Twelve services (5%) had or were developing treatment guidelines for heart failure: five were end of life pathways, three covered breathlessness management and three were symptom control guidelines. Some 137 services (59%) described local collaborative initiatives between SPC, heart failure services and primary care, such as mutual education, joint working and working groups. A number of models of joint workin